Objective To explore the differences in therapeutic efficacy and possible mechanism of different routes of administration of fullerene nanoparticles in the treatment of radiation retinopathy.Methods Eight-week-old adult male SD rats were randomly divided into blank group(Control),irradiation group(X-Ray),irradiation+vitreous cavity administration group[X+F(IVT)],irradiation+ocular surface administration group[X+F(OS)],and irradiation+intravenous administration group[X+F(IV)],with 5 rats in each group.The blank group was not treated,the irradiation groups exposed to X-ray irradiation to establish the model,and fullerenol nanoparticles were given to the treatment groups through different routes after irradiation.At 7,14,and 28 d after modeling,body weight and fundus changes were measured to evaluate drug safety,retinal optical coherence tomography(OCT)was used to observe the change in retinal tissue structure,and electroretinography(ERG)was applied for oscillatory potentials(OPs)to evaluate visual function.CD31 immunofluorescence staining was carried out to evaluate retinal endothelial vascular status,and in vivo imaging was utilized to evaluate the accumulation of fullerenol nanoparticles in the eyes.Results The growth curves of body weight demonstrated that fullerenol nanoparticles did not affect the growth and development of rats,with no statistical difference between the treatment groups and the control group.Irradiation resulted in a significant reduction in visual function,decreased amplitudes of a-wave and b-wave,and declined OPs(P<0.01),and significantly increased thicknesses of the ganglion cell layer(GCL)and the inner nuclear layer(INL)in the retinas,as evidenced by OCT(P<0.01),along with a notably absent presence of CD31-positive cells(P<0.01).Notably,the X+F(IVT)group obtained significantly improved visual function after intravitreal administration,effectively maintained thickness of the GCL and INL,and prevention against the loss of CD31-positive cells(P<0.01).However,no such effective results were observed in the irradiated groups receiving intravenous either ocular surface administration.In vivo imaging revealed that intravitreal administration maintained high ocular accumulation of fullerene for 96 h,while ocular surface administration sustained these concentrations for only 12 h.Intravenous administration,in contrast,only led to a predominant drug distribution in vascular-rich areas,but reduced ocular accumulation.Conclusion Fullerene nanoparticles possess a therapeutic effect on radiation retinopathy,and the intravitreal administration route demonstrates better efficacy than ocular surface and intravenous administration.